The Exon Theory of Genes hypothesizes that ancient genomes had an intron-exon structure that allowed for rapid diversification of protein structure and function through recombination of exons. While much research has investigated which modern exons correspond to structural units in proteins, and structural models have identified putative protein building blocks ubiquitous among non-homologous proteins, the principles that govern whether a polypeptide can be exchanged among different proteins remain unclear. We have developed an algorithm, called SCHEMA, to predict what elements, or schemata, of a protein can be swapped among homologous proteins without disrupting the folded structure. We propose a combination of biochemical and computational studies using SCHEMA and other novel algorithms, whose goals are to elucidate the rules governing non-disruptive recombination and evolution of novel functions by recombination. Our specific aims are to: 1) determine the SCHEMA-predicted threshold(s) of tolerable structural disruption upon homologous recombination for lactamases and cytochrome P450 monooxygenases; 2) develop novel algorithms for predicting efficient recombination fitness searches; 3) characterize the effectiveness of predicted search strategies through laboratory evolution of lactamases and cytochrome P450s; 4) optimize predictions of recombinant structural disruption; and 5) investigate if nonhomologous proteins can be recombined to generate folded proteins, using the algorithms to guide crossover locations. These studies should allow us to discover when homologous and non-homologous recombination conserves protein structure and expand our understanding of how evolution explores sequence, structural, and functional diversity. Furthermore, these studies should generate new tools for protein engineering by laboratory evolution, with biomedical applications in the development of new biomaterials, biosensors, catalysts, and protein-based therapeutics.